Mr. Ballmer, Show Us the Code
DigDuality writes "A new campaign, Showusthecode.com, requests every leader in the Linux world, and companies invested in Linux, to stand up and demand that Steve Ballmer show the world where Linux violates Microsoft's intellectual property. He has been making these claims since the Novell-Microsoft deal. If Microsoft answers this challenge — by May 1st — then Linux developers will be able to modify the code so that it remains 'free' software. If such infringing code doesn't exist, we will have called Microsoft's bluff. And if the campaign garners enough attention and if Steve Ballmer maintains silence, then the community and companies behind Linux can take the silence for the admission that it is."
Dude, just stop talking, you don't know what you're talking about, and it's embarrassing everyone reading this thread. The OP is basically right about everything he said, and you are similarly wrong about nearly everything you're asserting.
Your Ogg Vorbis argument is ridiculous, as patents are typically very specific in nature (as the OP said, they do not cover over-arching architectural design -- POSIX was a stupid argument, and so is this isomorphic audio codec argument). Furthermore, your moronic assertion that "if it used a different algorithm to derive that format it wouldn't infringe the [mp3] patent" assumes (for no reason) that the relevant patent in his hypothetical scenario wasn't one relating to the format itself (yes, this can be done, and is, frequently).
Then you go on to condescendingly explain what a patent is supposed to do, as if the OP doesn't know -- when it's clear to everyone that he has a much better notion of how patents work than you do.
As for your BS about SMB, who the hell cares whether IBM developed it or not? Did you hatch from a spore or something? Nearly all of Microsoft's technologies were developed by someone else originally -- which has absolutely no bearing whatsoever on whether or not Microsoft has added features or extensions that they've patented. Listen, Samba is not trying to interoperate with IBM's SMB implementation -- in fact, regardless of whether IBM developed SMB or not, the only SMB implementation that matters is Microsoft's, which has changed a lot over the years. Samba aims to be compatible with Microsoft, not IBM. In order to do this, it is exceedingly likely that Samba infringes on their patents. And, just for the record, most modern MS-based filesharing uses CIFS, not SMB (Samba does both). The former is much more of a Microsoft technology.
I also love how you essentially say "since MS and Samba co-operated at one point in the past" (an assertion I'm not even sure is true, but I'll let that slide because I'll admit that I don't know) "there can be no patent violations in Samba". I mean, holy shit! Are you daft? Do you honestly believe the world actually works this way? The only thing that results in indemnity from patent litigation is a patent license. "But your honor, MS helped me develop this piece of software, there couldn't be any patents in it aww gee shucks" is not valid legal recourse. In fact, this is one of the basic strategies involved in submarine patenting -- get your patented procedures, formats, and languages in widespread use however you can (often by publically releasing code in violation) and then, years later, when it's a standard, sue the pants off of whoever has money.
The OP is right on -- there's no code to be shown in a patent violations case. If they sue you for patent violations, the burden is on exactly one person -- YOU -- to show that your implementation of the algorithm is not in fact in violation of their patents. They don't have to find the code in question for you. It's a simple as "This feature of NTFS is patented by patent X and the Linux kernel implements NTFS in its entirety -- therefore it must be infringing. Pay us or we'll take you to court."
There is absolutely no question that Linux violates a huge number of patents. This is why, as the OP said before, software patents are evil.